Numerical Simulation of flow around the squareback of Ahmed body by PowerFlow.docx

Numerical Simulation of flow around the squareback of Ahmed body by PowerFlow.docx

《Numerical Simulation of flow around the squareback of Ahmed body by PowerFlow.docx》由会员分享，可在线阅读，更多相关《Numerical Simulation of flow around the squareback of Ahmed body by PowerFlow.docx（43页珍藏版）》请在冰豆网上搜索。

NumericalSimulationofflowaroundthesquarebackofAhmedbodybyPowerFlow

Studienarbeit（M.Sc.）

NumerischeStrömungssimulationumeinVollheckAhmedModellsmitPowerFlow

NumericalsimulationoftheflowaroundasquarebackAhmedmodelwithPowerFlow

von

cand.fmtXinleiMa

Matr.Nr.3101346

UniversitätStuttgart

InstitutfürVerbrennungsmotorenundKraftfahrwesen

LehrstuhlKraftfahrwesen

Prof.Dr.-Ing.J.Wiedemann

2017.8.20

EidesstattlicheErklärung

Hiermitversichereich,XinleiMa,dassichdievorliegendeArbeitbzw.diedarinmitmeinemNamengekennzeichnetenAnteileselbständigverfasstundnurdieangegebenenQuellenundHilfsmittelbenutzthabe.DabeihabeichallewörtlichodersinngemäßausanderenWerkenübernommenenAussagenalssolchegekenn-zeichnet.

DieArbeitistwedervollständignochinwesentlichenTeilenGegenstandeinesanderenPrüfungsverfahrensgewesen.FerneristsiewedervollständignochinTeilenbereitsveröffentlichtworden.DaselektronischeExemplarstimmtmitdenanderenExemplarenüberein.

Stuttgart,20.08.2017

XinleiMa

Declaration

HerebyIdeclare,XinleiMa,whichthepresentworkandthesharesmarkedwithmynamewerewrittenindependentlyandonlytheindicatedsourcesandtoolswereused.Thequotationsandreferenceshavebeendulyacknowledgedintheconcernedplaces.

Theworkhasnotbeenthesubjectofanyotherexaminationprocedure,eitherwhollyorinparts.Furthermore,ithasnotbeenpublishedeithercompletelyorinpart.Theelectroniccopyagreeswiththeothercopies.

Stuttgart,20.08.2017

XinleiMa

Content

AbstractIII

AbbreviationsIV

SymbolsV

ListofFiguresVII

ListofTablesIX

1Introduction1

1.1GeneralResearchBackgroundandImplication1

1.2ResearchMethod–ComputationalFluidDynamicsNumericalSimulation（CFD）2

1.3CurrentSituationofCFDAutomotiveResearch3

1.4TheResearchContent4

2CFDFundamentalTheory5

2.1CFDFundamentalTheory5

2.1.1FluidDynamicGoverningEquations5

2.1.2ABriefIntroductiontotheGridinCFD6

2.1.3NumericalSolutioninCFD7

2.1.4ABriefIntroductiontoTurbulence7

2.2TheProcessandCharacteristicsoftheOutflowFieldSimulation8

2.2.1TheProcessoftheOutflowFieldSimulation8

2.2.2TheCharacteristicsoftheOutflowFieldSimulation9

3TheSimulationontheAhmedBody10

3.1NumericalSimulationwithPowerFlow10

3.2AhmedBodyModel12

3.3SimulationEnviroment15

3.4GridGenerationandVRlevels16

4ResultandDiscussion18

4.1Streamlinesoftime-averagedvelocityintheWakeFlow18

4.2VelocityProfiles23

4.3TheInfluenceofResolutiononSimulationResults27

5Conclusions30

6Literature31

Appendix33

A.1ExamplesummaryinPowerFlow33

A.2CalculationcontentsinPowerFlow40

Abstract

Theautomobileindustryisdevelopingrapidly,thecarownershipisrisingsofast.Thegreatquantityofthevehiclesnotonlyconsumesthelargeamountofthepetroleumreserve,butalsotheenvironmentalpollutionproblemsaremuchmoreserious.Underthebackgroundoftheenergyconservationandtheenvironmentalprotection,thedemandofautomobileenergysavingandconsumptionisurgent.Theenergyconsumptionfromthevehicleislargelytoovercometheairresistanceduringthedriving,whichmakestheautomobileaerodynamicresearchbecomeoneofthehotspots.

Theapplicationofnumericalsimulationbasedoncomputationalfluiddynamics（CFD）isbecomingmoreandmorepopularinthestudyofautomobileflowfield.Comparingwiththewindtunneltest,thenumeriacalsimulationmethodhastheadvantagesoflowcostandshortcycle.

Theairresistancefromthevehicleislargelyduetothevorticesgeneratedbythetail.Therefore,thetailairflowundertheeffectivecontrollingbecomesthekeytothedragreduction.

InthisthesisthecommercialcomputationalfluiddynamicssoftwarePowerFlowisused,therewedotheresearchaboutthenumericalsimulationoftheflowfieldaroundtheAhmedbodyunderthedifferentaccuracyofthesimulationonthetailoftheAhmedbody,wecollectandcomparethedatatoexplorethepriciplesoftheairresistanceofthesquarebackonAhmedbody.Mesurethesteamlineofthetime-averagedvelocityandvelocityprofileatdifferentwakelocationwithdifferentresolutioncases.Findouttheinfluenceoftheresolutiononthesimulationwiththecomparisonwiththeexperimentdata.Summarizetheeffectofsimulationresultsunderdifferentaccuracy.

Abbreviations

CFD

ComputationalFluidDynamics

RSM

ReynoldsStressModel

DNS

DirectlyNumericalSimulationmethod

LBM

Lattice-Boltzmann-Method

NS-CFD

Navier-StokesComputationalFluidDynamicsmethods

UDDS

UrbanDynamometerDrivingSchedule

FDM

FiniteDifferenceMethod

FEM

FiniteElementMethod

FVM

FiniteVolumeMethod

CAD

ComputerAidedDesign

RANS

Reynolds-AveragedNavier-Stokessolver

VR

VariablesResolution

FeV

FinestregionofVoxels

FeS

FinestregionofSurfels

PIV

ParticleImageVelocimetry

LES

LargerEddySimulation

Symbols

Volumen

Density

Time

Thetotaldragcoefficientincludingfrictionalresistancecoefficientanddifferentialpressureresistancecoefficient

Thetotalfrictionalresistancecoefficient

Thefrontenddifferentialpressurecoefficient

Thedifferentialpressurecoefficientoftheverticalplaneofthetail

Thedifferentialpressurecoefficientofthetailslope

Free-streamvelocity

Taylormicroscale

Minimalcellsize

Reynoldsnumber

∆t

Onetimestepinseconds

Averaged-time

ElapsedtimethroughtheAhmedbodylength

y*

Thedistancefromthecentertothetoptrailingedge

ListofFigures

Figure1.1Theproportionofpneumaticresistancetototalresistance1

Figure2.1Thesketchmapsofthegrid7

Figure2.2Themeasuredvelocityatapointinaturbulentflow7

Figure2.3CFDBasicFlowDiagram8

Figure3.1Theelementsinthelattice10

Figure3.2TheoreticalApproachesofDIGITALPHYSICSandTraditionalCFD11

Figure3.3ThebasicdimensionsoftheAhmedmodel13

Figure3.4ThedragcoefficientsoftheAhmedbodyfrom0-40angle14

Figure3.5Theexperimentset-up15

Figure3.6ThemeshmodebyANSA16

Figure3.7DifferentVRlevelsaroundtheAhmedbodymodel17

Figure4.1ComparisonbetweenexperimentalPIVdataandthesimulationdatainthelongitudinalsymmetricalplane.（a）time-averagedPIV;（b）time-averagedLES;（c）time-averagedPowerFlow19-20

Figure4.2ThevorticeslocationsinPowerFlowsimulation21

Figure4.3Theboundarylayerdevelopedbetweenthevortices21

Figure4.4SurfacestreamlineshowingtheconvergentpointN22

Figure4.5Thevorticesweregeneratedaroundthesquareback22

Figure4.6Comparisonofthetime-averagedstreamwisevelocitycomponent,ufordifferentlocationsinwake.Shearlayerprofile（0.03H）downstreamofthetoptrailingedge.y∗=y+0.5H23

Figure4.7Comparisonprofileat（a）0.17Hdownstream（b）0.34Hdownstream（c）0.5Hdownstream（d）0.67Hdownstream（f）0.84Hdownstream24-26

Figure4.8Streamwisevelocitydistributionwirhfinestresolution27

Figure4.9（a）SteamwiseVelocityDistributionChangewiththedifferentgridsize（b）VelocityProfileDistributionChangewiththedifferentgridsize27-28

ListofTables

Table 3.1:

Numberoflatticesandcomputationaleffortassociatedwiththeresolution17

Table4.1:

Differentresolutionwithtotalsimulationtime28

1Introduction

1.1GeneralResearchBackgroundandImplication

Automotiveaerodynamicsisthesicencethatstudiestheinteractionbetweenthevehiclesandair.Asoneimportantperformanceofthevehicles,theaerodynamicscharacteristicsofthevehicleshasthegreatrelationwiththevehiclefueleconomy,controllingstability,safetyandcomfort.Especiallyundertheageoftheenergyconservationandtheenvironmentalprotection,thereisagreatvaluableapplicationinreducingfuelconsumptioninautomotiveaerodynamicsfield.

Thefigure1.1showstheproportionofpneumaticresistancecomparedtototalresistance.Atthevehiclespeed80km/h,thepneumaticresistanceisalmostequaltotherollingdynamicresistace;atthevehiclespeed150km/h,thepneumaticresistanceisequivalenttothe2~3timesoftherollingdynamicresistance[1].

Figure1.1Theproportionofpneumaticresistancetototalresistance

Thus,thereductionoftheaerodynamicdragissignificantforreducingthetotalresistanceofthevehicle,whichcangreatlyreducefuelconsumption.

Thepneumaticresistanceofthevehicleismainlydividedintotwoparts;pressureresistanceandfrictionresistance.TheGermanProfessorS.R.Ahmedresearchshowsthatforagroundvehicleshapebluntbodymodel-Ahmedbodymodel,thediffentialpressureresistanceisasmuchas85%ofthetotalresistance,therestisthefrictionresistance;onlythe9%ofthediffentialpressureresistanceisproducedfromthefront,mostofthediffentialpressureresistancearegeneratedbythetail[2].

Insummary,wehavethereasontobelievethatthekeyofthereductionoftheenergy-efficientvehicleistoreducethevehicle'saerodynamicdragandthecoreofthereductionofthepneumaticresistanceistosuppressthediffentialpressureresistancebytail.

1.2ResearchMethod–ComputationalFluidDynamicsNumericalSimulation（CFD）

Atpresent,ourresearchmethodsare:

theoreticalanalysis,roadtest,windtunneltest,computationalfluiddynamicsnumericalsimulation（CFD）.

Foralongperiodofyears,thewindtunneltesthasbeenamajortoolforevaluatingthevehicleaerodynamicperformance.However,inrecentyears,thedevelopmentofcomputerhardwareandsoftwaremakespossibleforthehigh-qualityautomoti